A regional, size-dependent, and causal effective medium model for Asian and Saharan mineral dust refractive index spectra

A model is developed for the refractive index spectra of desert mineral dust. This model is applicable regionally to both Asian and Saharan dust as the largest global aerosol sources. The capability of the model further aims at representing important local features through a subdivision into northern and southern Sahara, as well as western and eastern Asia. Machine learning techniques for accelerated literature data acquisition are presented. Available refractive index spectra for individual minerals and chemical species are combined based on the Bruggeman effective medium formula. A numerical procedure for effectively solving the resulting higher-order polynomial expression is presented. The present results of the effective refractive indices are validated through the Kramers-Kronig relation; in particular, a Hilbert transform is applied to the imaginary part of the refractive index spectra.